6278
viously12 in the model system IV. Initially two substituents, one electron donating (CH,) and one electron withdrawing (Br), were chosen for study for reasons of synthetic convenience. Tanida has previously determined the relative order of reactivity of esters IX-OBs, IV-OBs, and VIII-OBs to be 5.7 : 1.0:0.03, respectively, at 77.6”.12 If a simple steric effect is operative a constant additive factor for the exo cyclopropyl group should obtain and the ratio of relative reactivities for brosylates XI-OBs, VI-OBs, and X-OBs should be essentially identical with that found by Tanida.12 For an electronic effect such as a, however, one would anticipate substituent modulation of the degree of electron release from the cyclopropane edge bond with a resultant compression of the relative reactivity order XI : VI:X. The acetolysis rate ratios for compounds X-OBs* and XI-OBss (Table 11) relative to VI-OBs were found to be within experimental error the same as those reported by Tanida for the corresponding substituent effects in the model 7-benznorbornenyl series. From these initial results we conclude that the evidence is most consistent with a simple steric accelerative effect;13 however, we cannot at this time rule out steric electronic argument c due to uncertainties regarding the sensitivity of substituent probing to the latter effect. Acknowledgment. Financial support of this research by the National Science Foundation (Grant No. GP-27411 and GP-9412) is gratefully acknowledged. (12) H. Tanida, T. Tsuji, andH. Ishitobi, J . Amer. Chem. Soc., 86,4904 (1964); see also H. Tanida, Accounrs Chem. Res., 1,239:(1968). (13) In an attempt to account for the enhanced reactivity of exosyn-8-tricyclo[3.2.1.02.”]octyl brosylate (i) compared to its endo-syn isomer ii Haywood-Farmer and Pincock6coffered two possible explanations
11
1 a favorable stereoelectronic interaction of the exo cyclopropyl group with the incipient carbonium ion center that is formed during concerted Wagner-Meerwein rearrangement of the C-1-C-7 bond or, alternatively, steric acceleration by the proximate methylene group of the exo cyclopropyl ring. Our results suggest that it is the latter effect which accounts for most, if not all, of the rate enhancement of i over ii. (14) National Science Foundation Cooperative Fellow, 1966-1970.
Merle A. Battiste,* Paul F. Ranken,“ Robert Edelman Department of Chemistry, University of Florida Gainesville, Florida 32601 Received September 3, 1971
u-T Conjugation of Carbon-Metal Bonds. Dehydrometalation of Alkylmetal Compounds’
Sir: We wish to report an extraordinarily facile hydride abstraction from positions fi to carbon-metal bonds. ( I ) Supported by the Air Force Office of Scientific Research (Grant NO. AFOSR-69-1639A). (2) (a) Hydride abstraction from ethyl-transition metal compounds has been observedzb and attributed to *-complex formation. That the authors chose to offer this reactivity as a property exclusive to transition metalszc is testament to the misleading nature of the postulation of dorbital participation in such cations. Not only is the reactivity not confined to transition metals, but the reactivity is greater in certain
Journal of the American Chemical S o c i e t y
93:23
R2C-CRzmetalR,
I
+ PhsC+
--3
H Ph3CH
+ R2C=CR2 + metal+R,
(1)
Previous studies revealed a very large effect of the groups R,metalCH2- upon the vertical stabilization of positive carbon in such processes as eq 2.3 These
TCNE
TCNE.-
findings suggested that cations such as Me3SnCH2CH2+ or MeHgCH2CH2+,although primary, should be more stable than trityl To test this possibility we combined triphenylmethyl tetrafluoroborate with eth~ltrimethyltin~ in methylene chloride. The trityl cation spectrum disappeared almost immediately and triphenylmethane, ethylene, and Me8SnBF4 were produced in nearly quantitative yields.5 Similarly rapid reactions occurred with diPh8C+BFa-
+ MezEtSn -+ PhsCH
+ CHzCHz + MeaSn+BFd-
(3)
butylmercury, PhCH2CH2PbPh3,3b (PhCH2C(-Me)H)2Hg,3b and PhC(-Me)HSnMe,. However, PhCH2SnMe3,3b Me4Sn, (PhCMe2CH2)Hg,8 and di- I-apocamphylmercurys did not react, illustrating the importance of the presence and proper geometry of the hydrogen. The reaction obeys the rate expression
- d[Ph3C+1 = k2[Ph3C+][R,MCR2CHR2] dt
(4)
and shows a very large variation in rate with the nature of the metal in keeping with predictions based on u--?T conjugation effects.3b Rate data, shown in Table I, illustrate this and other effects. We have previously demonstrated3‘ the existence of a fi-alkylmercuricarbonium ion (mercurinium ion) as an intermediate in reaction 5 and the rearrangement of Me3SiCH2CD2+, observed by Cook, Eaborn, and W a l t ~ n ,clearly ~ demonstrates the existence of this /3 nontransition metal compounds such as EtaPb, etc. See ref 3a,b for further discussions of this matter, (b) M. L. H. Green and P. L. I. Nagy, J . Orgunomelal. Chem., 1, 5 8 (1963). (c) C. E. Coates, M. L. H. Green, and K. Wade, “Organometallic Chemistry,” Vol. 11, Methuen, London, 1968, p 216. (3) (a) J. C. Ware and T. G. Traylor, J. Amer. Chem. Soc., 89, 2304 (1967); (b) W. Hanstein, H. J. Berwin, and T. G. Traylor, ibid., 92, 829 (1970); (c) N. A. Clinton, R. S . Brown, and T. G. Traylor, ibid., 92, 5228 (1970); (d) W. Hanstein, H. J. Berwin, and T.G. Traylor, fbid., 92. 7476 (1970): (e) . W. Hanstein and T. G. Traylor, Terrahedron Letr., 4451 (1967). (4) Prepared by the method of Z. M. Manulkin, Zh. Obshch. Khim., 13, 46 (1943); Chem. Abstr., 38, 332 (1944). ( 5 ) Ethylene, identified by mass spectrometry, was evolved in 97% yield. Evaporation of the solvent followed by benzene extraction ofthe residue, washing, and drying afforded a 73 % yield of triphenylmethane. Trimethyltin fluoroboratea was identified by nmr. Similar results were obtained with the other organometallic compounds both in methylene chloride and in acetonitrile. I n neither solvent have we been able to observe the carbonium ions by nmr or to trap them by subsequent borohydride addition. However, Olah and Clifford’ have observed the RHgCH%CHa+ion (which we derive from EtzHg) in less nucleophilic solvents. Studies of reaction 1 in such solvents are in progress. (6) H. C. Clark and R. J. O’Brien, Znorg. Chem., 2, 1020 (1963). (7) G. A. Olah and P. R. Clifford, J . Amer. Chem. Soc., 93, 1261 (1971). (8) S . Winstein and T. G. Traylor, ibid., 78, 2597 (1956). Other compounds are commercially available or were identified by spectra and analyses, (9) M. A. Cook, C. Eaborn, and D. R. M. Walton, J . Organomefal. Chem., 24, 301 (1970).
1 November 17, 1971
I ,
I
6279 Table I. Rates of Reaction of Organometallic Compounds with PhaC+BF4- in Acetonitrile a t 29.8" ~
~~
~~~~
kz, M
Compound ~
SK-'
kz' per hydrogen
~~~
0.413 6 . 8 X lO-' 4.8 x 1 . 0 X 10-2(25') 5.9 0.9 0.115 0.0956
Cycloheptatriene Et4Si Et4Ge Et4Sn Et 4Pb EtzHg (MeaSnCHz)2CHzb MeaSnCH2CHzPh
0.21
5.7 x 10-8 4 x 10-6 8 . 3 X lo-' 0.49 0.15 0.058 0.048
0 Rates measured spectrophotometrically as pseudo-first-order disappearance of PhC+BF4- except for those of E t 8 and Et4Ge which were measured as zero-order rates by glc observation of ethylene. The product of this reaction is allyltrimethyltin identified by nmr spectroscopy.
silicon-substituted carbonium ion. We have strong evidence that reaction 5 is accelerated by vertical
r
7
6+
~ t a b i l i z a t i o n ~to~ produce the geometry indicated. These and the observation (Table I) that a second CH2SnMe, accelerates the rate more than does a phenyl grouplo make it very probable that these reactions proceed through the carbonium ion as shown in eq 6 . ' 3 - l 5 R&HCR2metal
1
Ph3C
+ Ph3C+
J
--+
RzC=CR2
+
+metal
+J
H
A plot of log kp'against the charge-transfer frequencies of Ph,MCH2Ph-tetracyanoethylene (eq 2) in Figure 1 shows that the vertical stabilization by R,MCH2(10) A second neighboring group can accelerate the rate by only a factor of two" whereas there is no limit on the acceleration available from a second vertical stabilizing group. ' 2 Previous resultsadshow that -CHzSnMe3 inductive effect is like CHI so that inductive effects are also excluded. (11) See C. J. Lancelot and P. v. R. Schleyer (J. Amer. Chem. Soc., 91, 4296 (1969)) for a discussion of the acceleration provided by two bridging groups. (12) H. Hart and P. A. Law, {bid., 86, 1957 (1964). (13) We do not exclude bridging. We only conclude that, in many of the cases we have studied, bridging makes negligible contribution to the stabilization and is therefore of little importance in OUT discussions of reaction rates or cation stabilizations. Since the energy difference between the symmetrical and unsymmetrical mercurinium ion is thought to be very small, the suggested" corner attack in oxymercuration of 7,7dimethylnorbornene could still partake of the same stabilization of the mercurinium ion as do reactions with the less-hindered norbornene. This interpretation is different from that of Brown and Kawakami" who consider the norbornene mercuration intermediate to be like the norbornyl cation obtained by protonation of norbornene. (14) H . C. Brown and J. H. Kawakami, J . Amer. Chem. SOC.,92, 201 (1970). (15) The existence of +CHKHzSiMea is clearly demonstrated,g and our evidence favors mechanism 6 for other group IV organometallic compounds. However, one-step dehydrometalation is expected for very reactive compounds such as EtaB-, etc. Studies of these compounds will be reported later.
Figure 1. Plot of log k2' for reaction of PhaC+with Et,M (reaction 1) against charge-transfer frequencies of Ph,MCHZPh with tetracyanoethylene.3brd See k2' per hydrogen i n Table I.
is accurately reflected in the rates of hydride abstraction. This is an additional implication that the principal driving force for hydride abstraction from Et,M is vertical stabilization of an intermediate carbonium ion by Et,MCH2-.1't's This reaction offers an attractive alternative to solvolysis as a way of studying stable cations. We are extending this study to other u--a conjugated cations.3b (16) 0.A. Reutov, E. V. Uglova, V. C. Makhaev, and V. S . Petrosyan, Zh. Org. Khim., 2153 (1970), proposed a cyclic elimination in the reaction of diisobutylmercury with (C6Ha)aCBr by analogy to the cyclic mechanism generally written for reduction of Grignard reagents. However, such cis elimination is not favored in reaction 117 and it might be incorrect for Grignard reductions. The stereochemistries of these reactions are under investigation. (17) Unpublished work of M. H. Hall. (18) The nmr results of Olah and Clifford' are consistent with and interpreted in terms of a bridged mercurinium ion. However, our previous evidence3- for a mercurinium ion as a reaction intermediate indicated it to be u--?Tconjugated and not bridged.3e The structure of bridged ions in nonnucleophilic solvents is, in this case, clearly not an accurate description of the transition state leading to the ion. Recent molecular orbital calculations on both CHaCHKHz+ 19 and RHgCH1CH+CHaZO seem to confirm our postulate that V-T conjugation is in general much more important than bridging or fragmentation. (19) L. Radom, J. A. Pople, V. Buss, and P. v. R. Schleyer, J . Amer. Chem. Soc., 93, 1813 (1971). (20) R. S. Bach, private communication. (21) On leave of absence from Institute Ruder B6skovic, Zagreb, Yugoslavia.
Jogko M. Jerkunica,zl T. G. Traylor* Chemistry Department, Reoelle College University of California, San Diego L a Jolla, California 92037 Receioed February 13, 1971
Single-Crystal X-Ray Structures of Chemotherapeutic Agents. V. The Structures of 5-Diazo-6-methoxy-6-hydrouraciland 2 '-Deoxy-5-diazo-6-hydro-06,5'-cyclouridine Hemihydrate Sir: The 5-diazo derivatives of uracil and uridine have been shown to possess significant carcinostatic activity. 1 Communications to the Editor
